Constructions to-to i N, v tom-to i N, v tom-to i ves' N: a corpus-based study

The present study aims to verify a hypothesis on the nature of lexical co-occurrence. According to the traditional view word combination can be free or fixed. The meaning of free word combination is derived from the meanings of its elements according to grammatical and semantic rules. Fixed expressions are non-compositional; i.e., their meaning is not derived, or not fully derived, from the meaning of their constituent parts. With the appearance of large text corpora and the development of new approaches in linguistics such as Construction Grammar (CxG), it has become possible to study the frequency and hence the cognitive entrenchment of word combinations by empirical methods, particularly quantitative analysis. We are going to show that fixedness of word combinations is not necessarily connected with non-compositionality. Many constructions formed in accordance with the rules governing the co-occurrence of their elements can nevertheless be retained in memory as separate units. Using large text corpora for the empirical data we are going to analyze the types of construction to-to i N (that’s the N), v tom-to i N (that’s the N) and v tom-to i ves' N (that’s just the whole N) with similar semantics which are realized in a wide variety of tokens. We will demonstrate that some tokens of the constructions can be so frequent that they can be considered to be cognitively entrenched units and are preserved in memory as separate units of the language. Such units should be described as separate items of the lexicon. We are also going to identify some regularities of distribution of fillers of the slot N in these constructions and formulate rules governing this distribution.